How Crystal Predictions Work

Crystal profiles combine AI-based analysis with traditional human-based personality assessments and evaluations. In general, accuracy increases when more data is available to analyze (via public data sources and user input).

With AI-based analysis alone, Crystal consistently performs at a ~85% accuracy rate, according to self and peer-reporting. For profiles that also include user personality assessments (we refer to these as "verified profiles"), accuracy improves to ~95%.

Crystal partners with GDPR-compliant data providers to analyze publicly available personality signals from social networks and other web pages.

These signals include, but are not limited to: job titles, industries, work experiences, skills, interests, recommendations, awards, certifications, courses, organizations, professional affiliations, projects, publications, summaries, and other fields associated with a person's public presence.

Crystal does not use any internal or private data, such as email content, non-public social media posts, non-public website content, financial records, or other sensitive personal information, for its analysis. 

DISC is a popular, scientifically-validated behavioral framework that’s based on a four-factor model of personality, meaning that it classifies four different categories based on observed behavioral patterns: Dominance, Influence, Steadiness, and Conscientiousness. Read more about DISC here. 

Yes, Crystal is GDPR compliant and designed with data privacy as its first principle. Read more about our GDPR compliance and privacy practices here. 

Crystal also follows Google Chrome Web Store’s security best practices for every release and update. Our extension is consistently recognized with the “Featured” badge within the Web Store which means our extension follows Google’s technical best practices and meets a high standard of user experience and design.